Journal of Behavior Therapy and Experimental Psychiatry 36 (2005) 240–253 Schema modes and childhood abuse in borderline and antisocial personality disorders Jill Lobbestael , Arnoud Arntz, Simkje Sieswerda Department of Medical, Clinical and Experimental Psychology, Maastricht University, PO Box 616, 6200 MD Maastricht, The Netherlands Abstract Complex personality disorders (PDs) have been hypothesized to be characterized by alternating states of thinking, feeling and behavior, the so-called schema modes (Young, Klosko, & Weishaar (2003). Schema therapy: A practioner’s guide. New York: Guilford). The present study tested the applicability of this model to borderline personality disorders (BPD) and antisocial personality disorders (APD), and related it to a presumed common etiological factor, childhood trauma. Sixteen patients with BPD, 16 patients with APD and 16 nonpatient controls (all 50% of both sexes) completed a Schema Mode Questionnaire assessing cognitions, feelings and behaviors characteristic of six schema modes. Participants were interviewed to retrace abusive sexual, physical and emotional events before the age of 18. BPD as well as APD participants were characterized by four maladaptive modes (Detached Protector, Punitive Parent, Abandoned/Abused Child and Angry Child). APD displayed most characteristics of the Bully/Attack mode, though not significantly different from BPD. The Healthy Adult mode was of low presence in BPD and of high presence in APD and the nonpatients. Frequency and severity of the three kinds of abuse were equally high in both PD groups, and significantly higher than in nonpatients. r 2005 Elsevier Ltd. All rights reserved. Keywords: Borderline personality disorder; Antisocial personality disorder; Schema modes; Cognitive therapy; Schema focused therapy; Childhood abuse ARTICLE IN PRESS www.elsevier.com/locate/jbtep 0005-7916/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbtep.2005.05.006 Corresponding author. Tel.: +31 43 388 1611; fax: +31 43 388 4155. E-mail address: [email protected] (J. Lobbestael).
Schema modes and childhood abuse in borderline and antisocial personality disorders
Oct 15, 2022
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
doi:10.1016/j.jbtep.2005.05.0060005-7916/$ -
doi:10.1016/j
Schema modes and childhood abuse in borderline and antisocial personality disorders
Jill Lobbestael, Arnoud Arntz, Simkje Sieswerda
Department of Medical, Clinical and Experimental Psychology, Maastricht University,
PO Box 616, 6200 MD Maastricht, The Netherlands
Abstract
Complex personality disorders (PDs) have been hypothesized to be characterized by
alternating states of thinking, feeling and behavior, the so-called schema modes (Young,
Klosko, & Weishaar (2003). Schema therapy: A practioner’s guide. New York: Guilford). The
present study tested the applicability of this model to borderline personality disorders (BPD)
and antisocial personality disorders (APD), and related it to a presumed common etiological
factor, childhood trauma. Sixteen patients with BPD, 16 patients with APD and 16 nonpatient
controls (all 50% of both sexes) completed a Schema Mode Questionnaire assessing
cognitions, feelings and behaviors characteristic of six schema modes. Participants were
interviewed to retrace abusive sexual, physical and emotional events before the age of 18. BPD
as well as APD participants were characterized by four maladaptive modes (Detached
Protector, Punitive Parent, Abandoned/Abused Child and Angry Child). APD displayed most
characteristics of the Bully/Attack mode, though not significantly different from BPD. The
Healthy Adult mode was of low presence in BPD and of high presence in APD and the
nonpatients. Frequency and severity of the three kinds of abuse were equally high in both PD
groups, and significantly higher than in nonpatients.
r 2005 Elsevier Ltd. All rights reserved.
Keywords: Borderline personality disorder; Antisocial personality disorder; Schema modes; Cognitive
therapy; Schema focused therapy; Childhood abuse
see front matter r 2005 Elsevier Ltd. All rights reserved.
.jbtep.2005.05.006
nding author. Tel.: +3143 388 1611; fax: +31 43 388 4155.
dress: [email protected] (J. Lobbestael).
ARTICLE IN PRESS
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253 241
1. Introduction
Recent insights have lead to the view that complex personality disorder (PD) are not characterized by one set of pathogenic schemas, but by different sets that can be activated in alternation. Young for instance, has proposed schema modes as relatively independent organized patterns of thinking, feeling and behaving that underlie the different states of severe PD patients (Young, Klosko, & Weishaar, 2003). In Young’s view borderline personality disorder (BPD) and antisocial personality disorder (APD) patients are characterized by various pathogenic schema modes. They are assumed to suddenly flip from one mode into an-other, especially in reaction to environmental changes caused by important events. Young hypothesized that four modes are central to BPD: the Detached Protector, the Angry and Impulsive Child, the Abandoned Child (in following with the second author in order to emphasize the central role of abuse, this mode will be further referred to as the Abandoned and Abused Child (Arntz & Bogels, 2000)) and the Punitive Parent. There also is a Healthy Adult mode, however due to extreme psychopathology of these patients it is assumed to be of low presence. Young’s schema-mode model is the basis of his schema therapy for severe PD, an increasingly popular therapeutic approach of which the effectivity is high (Giesen-Bloo et al., 2005; Nordahl & Nysæter, 2005).
When patients find themselves in the Abandoned and Abused Child mode, they feel the enormous pain and fear of abandonment caused by their abusive history which expresses itself in depressive, fearful, desperate, and inferiority feelings. This mode can be evoked by (perceptions of) (threatening) abandonment and abuse. Sometimes the patient becomes rebellious against the (supposed) injustice (s)he had experienced; this elicits the state of the Angry and Impulsive Child in which all bottled up aggressive feelings discharge so that anger, manipulation and greed are acted out. The evocation of these two child-modes usually leads to activation of self- punishing moral rules, mostly the direct internalizations of the punishing behavior of one of the caregivers, accounting for the symbolic mode name of the Punishing Parent. In this mode, the patient is afraid (s)he did something wrong, sees him/ herself as evil and worthless because of feelings and desires that are (threatened to be) activated. As a consequence of this self-directed anger and hate develops and the patient will punish him/herself in one or another way. Most of the time however, the patient finds him/herself in the Detached Protector mode, where (s)he does not have to feel the emotions and pain caused by the three other modes. The patient does not feel emotions, is unaware of any problems and is seemingly compliant (Arntz & Bogels, 2000; Arntz & Kuipers, 1998; Young et al., 2003).
As to APD, Young states that beside the Healthy Adult mode and the four modes described above, there is a fifth pathological mode present in antisocials called the Bully and Attack mode. In this mode, the antisocial hurts other people to overcompensate or to cope with mistrust, abuse, deprivation and defectiveness (Young, 2002; Young et al., 2003).
A study by Arntz, Klokman, and Sieswerda (2005) investigated whether the four maladaptive schema modes are specific for BPD patients and whether BPD-relevant
ARTICLE IN PRESS
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253242
stress specifically increases one of the modes, the Detached Protector mode. The results indicated that BPD patients were indeed characterized by the modes. The stress induction induced negative emotions in all groups, but the BPD group was unique in that the Detached Protector increased significantly more than in cluster-C PD patients and nonpatient controls (all women).
The hypothesized similarity in schema modes of BPD and APD has not been studied yet. Nevertheless, at least two sets of empirical findings suggest that the overlap in schema modes may be considerable. First, it has been noted that there is a large overlap in symptomatic expression of the two PDs.
Several DSM-IV diagnostic criteria of BPD and APD are quite similar, such as affect instability, inappropriate, intense and poorly controlled expression of anger and impulsivity that is potentially self-damaging (Blais & Norman, 1997; Holdwick, Hilsenroth, Casttebuty, & Blais, 1998). Furthermore, epidemiological figures point to high percentages of overlap; between 10 and 47% of BPD patients also meet the criteria for APD and about 70% display antisocial behavior (Paris, 1997; Widiger, Frances, & Trull, 1987; Zanarini & Gunderson, 1997). Averaged over studies approximately 70% of the APD patients meet BPD criteria (Widiger & Corbitt, 1997). Furthermore, while the prevalence in the community of both BPD and APD is about 1–2%, the sex distribution for APD is 80% male and for BPD 80% female. This would seem to make them ‘mirror image’ disorders. The gender difference could account to a large degree for the differences between BPD and APD; the differences in behavior being aggressiveness in APD and victimization in BPD could be a reflection of gender differences between men who more frequently display externalizing behavior and women who show more internalizing behavior. It has even been suggested that the two actually concern one underlying disorder, which expresses itself in BPD with women and in APD with men (Hudziak et al., 1996; Widiger & Corbitt, 1997; Paris, 1997).
Second, there also seems to be a large overlap in etiological factors. Numerous studies over the past decade have pointed out the frequent occurrence of childhood trauma in patients with BPD. Between 1987 and 1992, eleven studies confirmed this high incidence of childhood trauma in borderline patients (Sabo, 1997).
There are also studies reporting a positive relation between childhood abuse and APD (Burgess, Hartman, & McCormack, 1987; Dodge, Pettit, Bates, & Valente, 1995; Dutton & Hart, 1992; Horwitz, Widom, McLaughlin, & White, 2001; Marchall & Cooke, 1999; Pollock et al., 1990; Wallen, 1992). The DSM-IV states that childhood abuse or neglect increases the probability of a conduct disorder evolving in APD (APA, 1995). Burgess et al. (1987) have suggested a link between sexual abuse in childhood and later externalizing social deviant behavior. Dutton and Hart (1992) decided from file research of 604 male prisoners that men who were abused in childhood are three times more at risk of displaying violent behavior compared to nonabused men.
Despite the fact that these data suggest a central role of childhood abuse in both BPD and APD, there are—to our knowledge—only two studies that directly compared the prevalence and severity of abuse between both groups. Zanarini and Gunderson (1997) found in both groups substantial figures of childhood neglect and
ARTICLE IN PRESS
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253 243
abuse, although verbal abuse and emotional withdrawal were reported by a significantly higher percentage of the BPD group compared to the APD group. A study by Herman, Perry, and van der Kolk (1989) found that BPD patients gave significantly higher reports of physical, sexual and witnessing violence traumas than patients with borderline traits and persons with no borderline diagnoses. No association was found for APD and trauma. However, this study did not concern a systematic comparison between both groups, instead BPD patients were compared with a group of persons with borderline traits and with a mixed nonborderline control group with schizotypal PD (N ¼ 6), APD (N ¼ 6) and bipolar II affective disorder (N ¼ 11) (Herman et al., 1989).
The aims of the present study were twofold. Firstly, to assess and compare the presence of the hypothesized schema modes in borderlines, antisocials and nonpatient controls. Secondly, the direct comparison of childhood abuse history in the three groups. In this study, gender was equally divided within both groups so that the probability to detect disorder-specific results is increased. This is of particular interest since gender plays an important role in the prevalence of abuse and the coping behavior of abused persons; girls are at two to three times greater risk for sexual victimization and women more often internalize the anger accompanying abuse, while men more often show an externalizing coping style (Carmen, Rieker, and Mills, 1984).
2. Method
2.1. Subjects
Sixteen patients with BPD, 16 APD patients and 16 nonpatients controls were included in this study. Gender was evenly distributed within the groups by planned stratification, so each group consisted of eight men and eight women. Patients were recruited in Belgium from three mental hospitals (OPZ Rekem, Medisch Centrum St-Jozef in Bilzen and Psychiatrisch Centrum Ziekeren in St-Truiden), a community mental health service (CCG Hasselt) and correctional institutions in Brugge, Gent and Antwerp. Normals were mostly hospital staff. The study obtained institutional Human Studies approval.
All subjects were screened with SCID-I (modules A–D) and SCID-II interviews. To be included, subjects had to be between 18 and 50 years of age, and of normal intelligence (IQ480). Patients were admitted to the BPD group when they met DSM-IV criteria for BPD and not more than two APD criteria. APD patients had to meet DSM-IV criteria for APD and not more than two BPD criteria. Exclusion from the study occurred if patients met the criteria of a psychotic or bipolar disorder. Exclusion criteria for normal subjects were axis I or II disorders, and two or more BPD or APD criteria.
No between-group differences were found on age and intelligence. Mean age of the total sample was 30.9 years (BPD: 31.4; APD: 31.1; nonpatients: 30.2), ages ranging in each group from 19 to 46 years. There was no difference between the groups
ARTICLE IN PRESS
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253244
concerning mean intelligence (BPD: 100.1; APD: 105.9; nonpatients: 107.9). Neither did the clinical groups differ significantly in the presence of mood disorders (BPD: 62.5%; APD: 31.3%, N ¼ 16, w2ð1Þ ¼ 3:14, p ¼ 0:077) or the mean number of axis II disorders (BPD: 1.88; APD: 1.27, Mann-Whitney corrected Z ¼ 1:60, p ¼ 0:11).
The APD patients were significantly lower educated than the control group (Kruskal-Wallis: w2½2;N ¼ 48 ¼ 10:57, p ¼ 0:005), and a higher percentage of the patients were single (Chi-square: w2½2;N ¼ 48 ¼ 6:10, p ¼ 0:047). The analyses were not corrected for these two variables, because it was reasoned that they were inherent to BPD and APD.
2.2. Procedure
Subjects were individually seen at one of the institutions or prisons in Belgium between February and August 2002. At the start of the research procedure, informed consent was obtained. Participants were interviewed with both SCIDs and, if inclusion and exclusion criteria were met, with an interview for traumatic experiences. Then participants filled out the Schema Mode Questionnaire.
2.3. Materials
Dutch versions of the SCID-I and SCID-II were used to assess DSM-IV axis I diagnoses and personality pathology (First, Spitzer, Gibbon, Williams, & Benjamin, 1997; First, Spitzer, Gibbon, Williams, & Benjamin, 1994; van Groenestijn, Akkerhuis, Kupka, Schneider, & Nolen, 1999; Weertman, Arntz, & Kerkhofs, 2000). Good factorial validity and good interrater reliability of the Dutch SCID-II have been demonstrated in other studies (Arntz, 1999; Weertman, Arntz, Dreesen, van Velzen, & Vertommen, 2003).
The Schema Mode Questionnaire was administered to assess the 6 schema modes under study. This questionnaire is largely based on the Schema Mode Questionnaire developed by Arntz et al. (2005) which assesses the presence of five modes i.e. the Detached Protector (e.g. ‘It is best to keep a distance from other people’, ‘I feel empty’), the Angry Child (e.g. ‘I have to ventilate my feelings and work them off’, ‘I directly satisfy my needs’), the Abandoned and Abused Child (e.g. ‘I am helpless and powerless’, ‘I ask for reassurance’), the Punitive Parent (e.g. ‘I am bad and deserve punishment’, ‘I feel guilty’) and the Healthy Adult mode (e.g. ‘I am worthwhile’, ‘I feel good’). Based on suggestions by Young (personal communica- tion), McGinn and Young (1996), Beck and Freeman (1990), Arntz and Kuipers (1998) and clinical observations, this questionnaire was supplemented by cognitions, emotions and behaviors characteristic of the Bully and Attack mode (e.g. ‘Attack is the best defence’, ‘I humiliate others’). The final Schema Mode Questionnaire consisted of seven items on cognitions,’ five on emotions and five on behavior for each mode. Items were randomized within each category, resulting in a three-part questionnaire. Participants were instructed to rate the degree in which they generally believed in the stated cognitions, experienced the feelings described and engaged in the behavior on 100mm VASs.
ARTICLE IN PRESS
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253 245
To assess for childhood abuse, an interview for traumatic events was used (Bossche, Kremers, Sieswerda, & Arntz, 1999). This interview retraces whether participants experienced certain abusive sexual, physical or emotional events before the age of 18. It specifies the actions, frequency, perpetrator(s), and the age of and the impact on the victims. The interview has predetermined answer categories and results in composite scores for sexual, physical and emotional abuse separately. The higher the composite score is, the higher are frequency and/or severity of abuse. These abuse scores were constructed out of the closeness of the perpetuators, the number of perpetrators, age-level at time of abuse (the younger the subject, the higher the score), duration (the longer the duration, the higher the score) and severity of what had happened. Internal consistencies of the subscales assessed with the Cronbach alpha proved excellent in the present sample: sexual abuse 0.82, physical abuse 0.91 and emotional abuse 0.90. To create a composite abuse score, Z-scores for each of the three types of abuse were computed and averaged.
3. Results
3.1. Schema modes in BPD and APD patients
The reliability of the Schema Mode Questionnaire was analysed. Five of the 102 items did not contribute to the internal consistencies of the subscales they were hypothesized to belong to. After elimination of these items, Chronbach’s alpha coefficients showed excellent internal consistencies (see Table 1).
Fig. 1 depicts the mean scores of the groups on the six schema modes. Group differences were tested by means of MANOVA and Bonferroni corrected pair-wise comparisons. A multivariate test indicates a highly significant group effect, FHotð12; 78Þ ¼ 18:07, po0:001. Univariate tests revealed significant group effects on all subscales, F (2, 45)45.59, po.007. The groups’ means and standard deviations and contrasts between groups are presented in Table 2.
The BPD group scored significantly higher on the four BPD-related schema modes, and significantly lower on the Healthy Adult mode than the APD and the control group.
Table 1
Internal consistencies of the schema mode subscales as assessed with the Schema Mode Questionnaire
Mode Internal consistency
Detached Protector 0.93
Angry Child 0.87
Punishing Parent 0.91
ARTICLE IN PRESS
BPD
APD
NPCo
Fig. 1. Schema mode ratings by the three groups. DP—Detached Protector; AnCh—Angry Child;
AACh—Abandoned and Abused Child; PP—Punishing; HA—Healthy Adult; BA—Bully and Attack.
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253246
Although borderlines tended to score higher on the Bully and Attack mode than the nonpatients, this difference did not reach significance. In turn, antisocials also scored significantly higher on the four BPD-related schema modes than the control group. However, antisocials had lower scores on these modes than borderlines did. The Bully and Attack mode was significantly higher present in the APD group than in the normal control group, but the two PD groups did not differ significantly in that mode. APD patients scored significantly higher than the BPD group on the Healthy Adult mode. In fact, the presence of this mode did not differ significantly between the APD and control group.
The influence of gender on the mean scores of the modes was also analysed. A multivariate test revealed a gender effect, FHotð6; 37Þ ¼ 2:79, p ¼ 0:024. Univariate tests indicated that only the Bully and Attack mode was significantly stronger in men than in women, F ð5; 42Þ ¼ 4:48, p ¼ 0:04. None of the modes showed a significant interaction between group and gender, F ð5; 42Þ40:12, p40.21.
To summarize, the modes of the Detached Protector, the Angry Child, the Abandoned and Abused Child and the Punitive parent are indeed, as hypothesized, characteristic of BPD patients and also, but in lower degree, of APD patients. The Bully and Attack mode appeared specific for the APD group, but the difference between APD and BPD failed to reach significance. The Healthy Adult mode was of low presence in the borderlines, while the antisocials reported this mode equally high as the nonpatients.
3.2. Childhood abuse
Fig. 2 demonstrates the mean composite scores of severity of sexual, physical and emotional abuse before the age of 18. A multivariate test indicated a highly
ARTICLE IN PRESS
Mean, standard deviation and contrasts between groups of the modes
Contrastij mi sdi mj sdj t P
Detached Protector
Angry Child
Abandoned and Abused Child
Punishing Parent
Healthy Adult
Attack and Bully
BPD–APD 24.83 16.76 32.77 17.05 1.54 0.32
BPD–NPCo 24.83 16.76 15.54 8.18 1.80 0.21
APD–NPCo 32.77 17.05 15.54 8.18 3.34 0.007
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253 247
significant group effect, FHotð6; 84Þ ¼ 2:31, po0:001. Univariate tests also revealed significant group effects on all subscales, F ð2; 45Þ417:02, p o0.001. The groups with borderline and APD reported significantly higher rates of the three kinds of abuse than the nonpatient group (see Table 3).
Although BPD had higher sexual abuse scores than APD, whereas APD had higher physical abuse scores than BPD, these differences did not reach significance. Standardized z total scores of abuse were also not significantly higher amongst borderlines than amongst antisocials, which indicates that the prevalence and severity of abuse did not differ between the two groups.
Abuse data were analyzed more in detail concerning duration of the abuse, the number of perpetrators and abuse actions and the age-level at time of abuse. Inspection of these data showed borderlines experienced a higher number of sexual abuse actions compared to antisocials (means for BPD: 3; APD: 1.50).…
doi:10.1016/j
Schema modes and childhood abuse in borderline and antisocial personality disorders
Jill Lobbestael, Arnoud Arntz, Simkje Sieswerda
Department of Medical, Clinical and Experimental Psychology, Maastricht University,
PO Box 616, 6200 MD Maastricht, The Netherlands
Abstract
Complex personality disorders (PDs) have been hypothesized to be characterized by
alternating states of thinking, feeling and behavior, the so-called schema modes (Young,
Klosko, & Weishaar (2003). Schema therapy: A practioner’s guide. New York: Guilford). The
present study tested the applicability of this model to borderline personality disorders (BPD)
and antisocial personality disorders (APD), and related it to a presumed common etiological
factor, childhood trauma. Sixteen patients with BPD, 16 patients with APD and 16 nonpatient
controls (all 50% of both sexes) completed a Schema Mode Questionnaire assessing
cognitions, feelings and behaviors characteristic of six schema modes. Participants were
interviewed to retrace abusive sexual, physical and emotional events before the age of 18. BPD
as well as APD participants were characterized by four maladaptive modes (Detached
Protector, Punitive Parent, Abandoned/Abused Child and Angry Child). APD displayed most
characteristics of the Bully/Attack mode, though not significantly different from BPD. The
Healthy Adult mode was of low presence in BPD and of high presence in APD and the
nonpatients. Frequency and severity of the three kinds of abuse were equally high in both PD
groups, and significantly higher than in nonpatients.
r 2005 Elsevier Ltd. All rights reserved.
Keywords: Borderline personality disorder; Antisocial personality disorder; Schema modes; Cognitive
therapy; Schema focused therapy; Childhood abuse
see front matter r 2005 Elsevier Ltd. All rights reserved.
.jbtep.2005.05.006
nding author. Tel.: +3143 388 1611; fax: +31 43 388 4155.
dress: [email protected] (J. Lobbestael).
ARTICLE IN PRESS
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253 241
1. Introduction
Recent insights have lead to the view that complex personality disorder (PD) are not characterized by one set of pathogenic schemas, but by different sets that can be activated in alternation. Young for instance, has proposed schema modes as relatively independent organized patterns of thinking, feeling and behaving that underlie the different states of severe PD patients (Young, Klosko, & Weishaar, 2003). In Young’s view borderline personality disorder (BPD) and antisocial personality disorder (APD) patients are characterized by various pathogenic schema modes. They are assumed to suddenly flip from one mode into an-other, especially in reaction to environmental changes caused by important events. Young hypothesized that four modes are central to BPD: the Detached Protector, the Angry and Impulsive Child, the Abandoned Child (in following with the second author in order to emphasize the central role of abuse, this mode will be further referred to as the Abandoned and Abused Child (Arntz & Bogels, 2000)) and the Punitive Parent. There also is a Healthy Adult mode, however due to extreme psychopathology of these patients it is assumed to be of low presence. Young’s schema-mode model is the basis of his schema therapy for severe PD, an increasingly popular therapeutic approach of which the effectivity is high (Giesen-Bloo et al., 2005; Nordahl & Nysæter, 2005).
When patients find themselves in the Abandoned and Abused Child mode, they feel the enormous pain and fear of abandonment caused by their abusive history which expresses itself in depressive, fearful, desperate, and inferiority feelings. This mode can be evoked by (perceptions of) (threatening) abandonment and abuse. Sometimes the patient becomes rebellious against the (supposed) injustice (s)he had experienced; this elicits the state of the Angry and Impulsive Child in which all bottled up aggressive feelings discharge so that anger, manipulation and greed are acted out. The evocation of these two child-modes usually leads to activation of self- punishing moral rules, mostly the direct internalizations of the punishing behavior of one of the caregivers, accounting for the symbolic mode name of the Punishing Parent. In this mode, the patient is afraid (s)he did something wrong, sees him/ herself as evil and worthless because of feelings and desires that are (threatened to be) activated. As a consequence of this self-directed anger and hate develops and the patient will punish him/herself in one or another way. Most of the time however, the patient finds him/herself in the Detached Protector mode, where (s)he does not have to feel the emotions and pain caused by the three other modes. The patient does not feel emotions, is unaware of any problems and is seemingly compliant (Arntz & Bogels, 2000; Arntz & Kuipers, 1998; Young et al., 2003).
As to APD, Young states that beside the Healthy Adult mode and the four modes described above, there is a fifth pathological mode present in antisocials called the Bully and Attack mode. In this mode, the antisocial hurts other people to overcompensate or to cope with mistrust, abuse, deprivation and defectiveness (Young, 2002; Young et al., 2003).
A study by Arntz, Klokman, and Sieswerda (2005) investigated whether the four maladaptive schema modes are specific for BPD patients and whether BPD-relevant
ARTICLE IN PRESS
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253242
stress specifically increases one of the modes, the Detached Protector mode. The results indicated that BPD patients were indeed characterized by the modes. The stress induction induced negative emotions in all groups, but the BPD group was unique in that the Detached Protector increased significantly more than in cluster-C PD patients and nonpatient controls (all women).
The hypothesized similarity in schema modes of BPD and APD has not been studied yet. Nevertheless, at least two sets of empirical findings suggest that the overlap in schema modes may be considerable. First, it has been noted that there is a large overlap in symptomatic expression of the two PDs.
Several DSM-IV diagnostic criteria of BPD and APD are quite similar, such as affect instability, inappropriate, intense and poorly controlled expression of anger and impulsivity that is potentially self-damaging (Blais & Norman, 1997; Holdwick, Hilsenroth, Casttebuty, & Blais, 1998). Furthermore, epidemiological figures point to high percentages of overlap; between 10 and 47% of BPD patients also meet the criteria for APD and about 70% display antisocial behavior (Paris, 1997; Widiger, Frances, & Trull, 1987; Zanarini & Gunderson, 1997). Averaged over studies approximately 70% of the APD patients meet BPD criteria (Widiger & Corbitt, 1997). Furthermore, while the prevalence in the community of both BPD and APD is about 1–2%, the sex distribution for APD is 80% male and for BPD 80% female. This would seem to make them ‘mirror image’ disorders. The gender difference could account to a large degree for the differences between BPD and APD; the differences in behavior being aggressiveness in APD and victimization in BPD could be a reflection of gender differences between men who more frequently display externalizing behavior and women who show more internalizing behavior. It has even been suggested that the two actually concern one underlying disorder, which expresses itself in BPD with women and in APD with men (Hudziak et al., 1996; Widiger & Corbitt, 1997; Paris, 1997).
Second, there also seems to be a large overlap in etiological factors. Numerous studies over the past decade have pointed out the frequent occurrence of childhood trauma in patients with BPD. Between 1987 and 1992, eleven studies confirmed this high incidence of childhood trauma in borderline patients (Sabo, 1997).
There are also studies reporting a positive relation between childhood abuse and APD (Burgess, Hartman, & McCormack, 1987; Dodge, Pettit, Bates, & Valente, 1995; Dutton & Hart, 1992; Horwitz, Widom, McLaughlin, & White, 2001; Marchall & Cooke, 1999; Pollock et al., 1990; Wallen, 1992). The DSM-IV states that childhood abuse or neglect increases the probability of a conduct disorder evolving in APD (APA, 1995). Burgess et al. (1987) have suggested a link between sexual abuse in childhood and later externalizing social deviant behavior. Dutton and Hart (1992) decided from file research of 604 male prisoners that men who were abused in childhood are three times more at risk of displaying violent behavior compared to nonabused men.
Despite the fact that these data suggest a central role of childhood abuse in both BPD and APD, there are—to our knowledge—only two studies that directly compared the prevalence and severity of abuse between both groups. Zanarini and Gunderson (1997) found in both groups substantial figures of childhood neglect and
ARTICLE IN PRESS
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253 243
abuse, although verbal abuse and emotional withdrawal were reported by a significantly higher percentage of the BPD group compared to the APD group. A study by Herman, Perry, and van der Kolk (1989) found that BPD patients gave significantly higher reports of physical, sexual and witnessing violence traumas than patients with borderline traits and persons with no borderline diagnoses. No association was found for APD and trauma. However, this study did not concern a systematic comparison between both groups, instead BPD patients were compared with a group of persons with borderline traits and with a mixed nonborderline control group with schizotypal PD (N ¼ 6), APD (N ¼ 6) and bipolar II affective disorder (N ¼ 11) (Herman et al., 1989).
The aims of the present study were twofold. Firstly, to assess and compare the presence of the hypothesized schema modes in borderlines, antisocials and nonpatient controls. Secondly, the direct comparison of childhood abuse history in the three groups. In this study, gender was equally divided within both groups so that the probability to detect disorder-specific results is increased. This is of particular interest since gender plays an important role in the prevalence of abuse and the coping behavior of abused persons; girls are at two to three times greater risk for sexual victimization and women more often internalize the anger accompanying abuse, while men more often show an externalizing coping style (Carmen, Rieker, and Mills, 1984).
2. Method
2.1. Subjects
Sixteen patients with BPD, 16 APD patients and 16 nonpatients controls were included in this study. Gender was evenly distributed within the groups by planned stratification, so each group consisted of eight men and eight women. Patients were recruited in Belgium from three mental hospitals (OPZ Rekem, Medisch Centrum St-Jozef in Bilzen and Psychiatrisch Centrum Ziekeren in St-Truiden), a community mental health service (CCG Hasselt) and correctional institutions in Brugge, Gent and Antwerp. Normals were mostly hospital staff. The study obtained institutional Human Studies approval.
All subjects were screened with SCID-I (modules A–D) and SCID-II interviews. To be included, subjects had to be between 18 and 50 years of age, and of normal intelligence (IQ480). Patients were admitted to the BPD group when they met DSM-IV criteria for BPD and not more than two APD criteria. APD patients had to meet DSM-IV criteria for APD and not more than two BPD criteria. Exclusion from the study occurred if patients met the criteria of a psychotic or bipolar disorder. Exclusion criteria for normal subjects were axis I or II disorders, and two or more BPD or APD criteria.
No between-group differences were found on age and intelligence. Mean age of the total sample was 30.9 years (BPD: 31.4; APD: 31.1; nonpatients: 30.2), ages ranging in each group from 19 to 46 years. There was no difference between the groups
ARTICLE IN PRESS
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253244
concerning mean intelligence (BPD: 100.1; APD: 105.9; nonpatients: 107.9). Neither did the clinical groups differ significantly in the presence of mood disorders (BPD: 62.5%; APD: 31.3%, N ¼ 16, w2ð1Þ ¼ 3:14, p ¼ 0:077) or the mean number of axis II disorders (BPD: 1.88; APD: 1.27, Mann-Whitney corrected Z ¼ 1:60, p ¼ 0:11).
The APD patients were significantly lower educated than the control group (Kruskal-Wallis: w2½2;N ¼ 48 ¼ 10:57, p ¼ 0:005), and a higher percentage of the patients were single (Chi-square: w2½2;N ¼ 48 ¼ 6:10, p ¼ 0:047). The analyses were not corrected for these two variables, because it was reasoned that they were inherent to BPD and APD.
2.2. Procedure
Subjects were individually seen at one of the institutions or prisons in Belgium between February and August 2002. At the start of the research procedure, informed consent was obtained. Participants were interviewed with both SCIDs and, if inclusion and exclusion criteria were met, with an interview for traumatic experiences. Then participants filled out the Schema Mode Questionnaire.
2.3. Materials
Dutch versions of the SCID-I and SCID-II were used to assess DSM-IV axis I diagnoses and personality pathology (First, Spitzer, Gibbon, Williams, & Benjamin, 1997; First, Spitzer, Gibbon, Williams, & Benjamin, 1994; van Groenestijn, Akkerhuis, Kupka, Schneider, & Nolen, 1999; Weertman, Arntz, & Kerkhofs, 2000). Good factorial validity and good interrater reliability of the Dutch SCID-II have been demonstrated in other studies (Arntz, 1999; Weertman, Arntz, Dreesen, van Velzen, & Vertommen, 2003).
The Schema Mode Questionnaire was administered to assess the 6 schema modes under study. This questionnaire is largely based on the Schema Mode Questionnaire developed by Arntz et al. (2005) which assesses the presence of five modes i.e. the Detached Protector (e.g. ‘It is best to keep a distance from other people’, ‘I feel empty’), the Angry Child (e.g. ‘I have to ventilate my feelings and work them off’, ‘I directly satisfy my needs’), the Abandoned and Abused Child (e.g. ‘I am helpless and powerless’, ‘I ask for reassurance’), the Punitive Parent (e.g. ‘I am bad and deserve punishment’, ‘I feel guilty’) and the Healthy Adult mode (e.g. ‘I am worthwhile’, ‘I feel good’). Based on suggestions by Young (personal communica- tion), McGinn and Young (1996), Beck and Freeman (1990), Arntz and Kuipers (1998) and clinical observations, this questionnaire was supplemented by cognitions, emotions and behaviors characteristic of the Bully and Attack mode (e.g. ‘Attack is the best defence’, ‘I humiliate others’). The final Schema Mode Questionnaire consisted of seven items on cognitions,’ five on emotions and five on behavior for each mode. Items were randomized within each category, resulting in a three-part questionnaire. Participants were instructed to rate the degree in which they generally believed in the stated cognitions, experienced the feelings described and engaged in the behavior on 100mm VASs.
ARTICLE IN PRESS
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253 245
To assess for childhood abuse, an interview for traumatic events was used (Bossche, Kremers, Sieswerda, & Arntz, 1999). This interview retraces whether participants experienced certain abusive sexual, physical or emotional events before the age of 18. It specifies the actions, frequency, perpetrator(s), and the age of and the impact on the victims. The interview has predetermined answer categories and results in composite scores for sexual, physical and emotional abuse separately. The higher the composite score is, the higher are frequency and/or severity of abuse. These abuse scores were constructed out of the closeness of the perpetuators, the number of perpetrators, age-level at time of abuse (the younger the subject, the higher the score), duration (the longer the duration, the higher the score) and severity of what had happened. Internal consistencies of the subscales assessed with the Cronbach alpha proved excellent in the present sample: sexual abuse 0.82, physical abuse 0.91 and emotional abuse 0.90. To create a composite abuse score, Z-scores for each of the three types of abuse were computed and averaged.
3. Results
3.1. Schema modes in BPD and APD patients
The reliability of the Schema Mode Questionnaire was analysed. Five of the 102 items did not contribute to the internal consistencies of the subscales they were hypothesized to belong to. After elimination of these items, Chronbach’s alpha coefficients showed excellent internal consistencies (see Table 1).
Fig. 1 depicts the mean scores of the groups on the six schema modes. Group differences were tested by means of MANOVA and Bonferroni corrected pair-wise comparisons. A multivariate test indicates a highly significant group effect, FHotð12; 78Þ ¼ 18:07, po0:001. Univariate tests revealed significant group effects on all subscales, F (2, 45)45.59, po.007. The groups’ means and standard deviations and contrasts between groups are presented in Table 2.
The BPD group scored significantly higher on the four BPD-related schema modes, and significantly lower on the Healthy Adult mode than the APD and the control group.
Table 1
Internal consistencies of the schema mode subscales as assessed with the Schema Mode Questionnaire
Mode Internal consistency
Detached Protector 0.93
Angry Child 0.87
Punishing Parent 0.91
ARTICLE IN PRESS
BPD
APD
NPCo
Fig. 1. Schema mode ratings by the three groups. DP—Detached Protector; AnCh—Angry Child;
AACh—Abandoned and Abused Child; PP—Punishing; HA—Healthy Adult; BA—Bully and Attack.
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253246
Although borderlines tended to score higher on the Bully and Attack mode than the nonpatients, this difference did not reach significance. In turn, antisocials also scored significantly higher on the four BPD-related schema modes than the control group. However, antisocials had lower scores on these modes than borderlines did. The Bully and Attack mode was significantly higher present in the APD group than in the normal control group, but the two PD groups did not differ significantly in that mode. APD patients scored significantly higher than the BPD group on the Healthy Adult mode. In fact, the presence of this mode did not differ significantly between the APD and control group.
The influence of gender on the mean scores of the modes was also analysed. A multivariate test revealed a gender effect, FHotð6; 37Þ ¼ 2:79, p ¼ 0:024. Univariate tests indicated that only the Bully and Attack mode was significantly stronger in men than in women, F ð5; 42Þ ¼ 4:48, p ¼ 0:04. None of the modes showed a significant interaction between group and gender, F ð5; 42Þ40:12, p40.21.
To summarize, the modes of the Detached Protector, the Angry Child, the Abandoned and Abused Child and the Punitive parent are indeed, as hypothesized, characteristic of BPD patients and also, but in lower degree, of APD patients. The Bully and Attack mode appeared specific for the APD group, but the difference between APD and BPD failed to reach significance. The Healthy Adult mode was of low presence in the borderlines, while the antisocials reported this mode equally high as the nonpatients.
3.2. Childhood abuse
Fig. 2 demonstrates the mean composite scores of severity of sexual, physical and emotional abuse before the age of 18. A multivariate test indicated a highly
ARTICLE IN PRESS
Mean, standard deviation and contrasts between groups of the modes
Contrastij mi sdi mj sdj t P
Detached Protector
Angry Child
Abandoned and Abused Child
Punishing Parent
Healthy Adult
Attack and Bully
BPD–APD 24.83 16.76 32.77 17.05 1.54 0.32
BPD–NPCo 24.83 16.76 15.54 8.18 1.80 0.21
APD–NPCo 32.77 17.05 15.54 8.18 3.34 0.007
J. Lobbestael et al. / J. Behav. Ther. & Exp. Psychiat. 36 (2005) 240–253 247
significant group effect, FHotð6; 84Þ ¼ 2:31, po0:001. Univariate tests also revealed significant group effects on all subscales, F ð2; 45Þ417:02, p o0.001. The groups with borderline and APD reported significantly higher rates of the three kinds of abuse than the nonpatient group (see Table 3).
Although BPD had higher sexual abuse scores than APD, whereas APD had higher physical abuse scores than BPD, these differences did not reach significance. Standardized z total scores of abuse were also not significantly higher amongst borderlines than amongst antisocials, which indicates that the prevalence and severity of abuse did not differ between the two groups.
Abuse data were analyzed more in detail concerning duration of the abuse, the number of perpetrators and abuse actions and the age-level at time of abuse. Inspection of these data showed borderlines experienced a higher number of sexual abuse actions compared to antisocials (means for BPD: 3; APD: 1.50).…
Related Documents
